Pre-clinical evaluation of a 213Bi-labeled 2556 antibody to HIV-1 gp41 glycoprotein in HIV-1 mouse models as a reagent for HIV eradication

Abstract

Background: Any strategy for curing HIV infection must include a method to eliminate viral-infected cells. Based on our earlier proof-of-principle results targeting HIV-1 infected cells with radiolabeled antibody (mAb) to gp41 viral antigen, we embarked on identifying a suitable candidate mAb for preclinical development.

Methodology/principal findings: Among the several human mAbs to gp41 tested, mAb 2556 was found to have high affinity, reactivity with multimeric forms of gp41 present on both the surface of virus particles and cells expressing HIV-1 Env, and recognition of a highly conserved epitope of gp41 shared by all HIV-1 subtypes. Also, mAb 2556 was the best in competition with HIV-1+ serum antibodies, which is an extremely important consideration for efficacy in the treatment of HIV patients. When radiolabeled with alpha-emitting radionuclide 213-Bismuth ((213)Bi) - (213)Bi-2556 efficiently and specifically killed ACH-2 human lymphocytes chronically infected with HIV-1, and HIV-1 infected human peripheral blood mononuclear cells (hPBMCs). The number of binding sites for (213)Bi-2556 on the surface of the infected cells was >10(6). The in vivo experiments were performed in two HIV-1 mouse models--splenic and intraperitoneal. In both models, the decrease in HIV-1 infected hPBMCs from the spleens and peritoneum, respectively, was dose-dependent with the most pronounced killing of hPBMCs observed in the 100 µCi (213)Bi-2556 group (P = 0.01). Measurement of the blood platelet counts and gross pathology of the treated mice demonstrated the lack of toxicity for (213)Bi-2556.

Conclusions/significance: We describe the preclinical development of a novel radiolabeled mAb reagent that could potentially be part of an HIV eradication strategy that is ready for translation into the clinic as the next step in its development. As viral antigens are very different from "self" human antigens - this approach promises high selectivity, increased efficacy and low toxicity, especially in comparison to immunotoxins.

Figure 1. Binding of mAb 2556 to 15-mer overlapping gp41 peptides.

The common sequence which 2556 reacts with the various overlapping peptides and corresponding OD values are labeled blue. MAb 2556 can bind to a linear epitope which is located in the region of the N-terminal cysteine (AA 591) of the immunodominant disulfide loop of gp41. The 2556 epitope contains two residues GC (bold and underlined) which are recognized by the mAb in all sequences (core epitope) and few residues in the vicinity which contribute to this epitope.

Figure 2. Characterization of 2556 mAb specificity and selectivity.

a) inhibition of binding of biotinylated anti-gp41 mAbs to JR-CSF transfected 293T cells by serum from HIV-1 infected individuals; b) binding of human anti-gp41 mAbs to recombinant gp41_MN protein; c) ELISA reactivity to recombinant gp41_MN of 2556 CHROMOS (made by ACE system) and 2556 (in house) derived from original NYU hybridoma. 246D is an anti-gp41 mAb and 1418 is a negative control.

Figure 3. Immunoreactivity of ²¹³Bi-2556 mAb for gp41 and determination of its K_a by Scatchard analysis.

a) gp41 ELISA of 2556 mAb radiolabeled with high specific activity ²¹³Bi (5 mCi/mg). Binding of ²¹³Bi-2556 to gp41 was compared to that of unlabeled 2556 mAb with (2556-CHXA″) or without (2556 standard) attached CHXA″ ligand. ²¹³Bi-2556 was stored with human serum albumin (HAS) radioprotector (2556-Bi/HAS) or without it (2556-Bi); b) Scatchard plot of ²¹³Bi-labeled 2556 and 246D mAbs binding to HIV-infected hPBMCs.

Figure 4. Targeting and killing of HIV-infected cells in vitro with ²¹³Bi-2556 mAb.

a) binding of “cold” 2556 to chronically-infected ACH-2 cells by flow cytometry. HIV-1-negative A3.01 cells were used as controls; b) killing of HIV-infected hPBMCs with ²¹³Bi-2556 mAb. 246D is an anti-gp41 mAb and 1418 is a negative control; c) killing of chronically-infected ACH-2 cells with ²¹³Bi-2556 mAb. Stimulated - PMA stimulated ACH-2 cells, unstimulated – unstimulated ACH-2 cells. The unlabeled antibodies were present in the same amounts as the radiolabeled antibodies.

Figure 5. RT-PCR data for two HIV-1 mouse models used in RIT experiments with ²¹³Bi-2556 mAb.

a) splenic model (mice given HIV-1 infected hPBMCs intrasplenically); b) peritoneal model (mice given non-infected hPBMCs followed by i.p. challenge with HIV). “Cold” 2556 – unlabeled 2556; 1418 – isotype-matching irrelevant control.

Figure 6. Peripheral blood platelet counts from SCID mice used in (a) splenic and (b) peritoneal HIV models and treated with ²¹³Bi-2556 and control mAbs.